I've been tampering with the asm a bit and had some luck, especially in scalar performance by merging the various stages into one function, however do note that it's gcc only (clang had some issues with rsi/rdi use on the manually inlined function and to work around it one has to move rsi or rdi to some xmm register and restore it before the function ends - which adds a few cycles)

https://github.com/Alex-GR/secp256k1/blob/master/src/scalar_4x64_impl.h
https://github.com/Alex-GR/secp256k1/blob/master/src/field_5x52_asm_impl.h

In 5x52 asm I found 2-3% by simply doing what the cpu scheduler should be doing and issuing together the adds + muls (the cpu integer unit typically has one add and one mul unit and ideally we want to be using them at the same time).

For example:

/* d += a3 * b1 */
    "movq 8(%%rbx),%%rax\n"
    "mulq %%r13\n"
    "addq %%rax,%%rcx\n"
    "adcq %%rdx,%%r15\n"
    /* d += a2 * b2 */
    "movq 16(%%rbx),%%rax\n" <=== this must be moved upwards
    "mulq %%r12\n"

becomes:

/* d += a3 * b1 */
    "movq 8(%%rbx),%%rax\n"
    "mulq %%r13\n"
    "addq %%rax,%%rcx\n"
    "movq 16(%%rbx),%%rax\n"
    "adcq %%rdx,%%r15\n"
    /* d += a2 * b2 */
    "mulq %%r12\n"

So the adcq+mulq are issued together to the mul and add units of the cpu respectively. (edit apparently the mul+add are in the simd unit, in the integer part it's just 3 integer units waiting to do parallel work of any kind - except load/store, which is 2 at a time). I'm baffled on why the cpu scheduler wasn't already doing this but then again I do have an older cpu (core2 quad / 45nm) to play with - it might not be an issue with modern ones.

Also, of the three temp variables used for temp storage in the field asm, some can be eliminated by rearranging the code a bit and thus reducing a couple memory accesses.

(ps. I don't claim the code is safe - I've only used it for benchmarks and the builtin test)

One question I have regarding secp256k1 is whether endomorphism is safe, and if yes, shouldn't it be enabled in bitcoin builds if it's faster (benchmarks show that it is)?

It depends on what you expect it to do. If you view the longer-term trend as increasing, then a temporary (mostly artificial) fall is a bear trap. If, on the other hand, expect the long-term trend to be decreasing, then a temporary (mostly artificial) rise is a bull trap.

I think if there was real selling pressure, we'd see this reflected in the order books. There's not too many BTCs on the upside. A few thousand coins take us back to 1000$. On the other hand, it takes multiple that to go down to 700. So, given the positioning and strength of the buys and sells on the order book, I'd say the market expects BTC prices to rise.

Interesting. Very thin on the sell side, pretty strong on the buy side. This is more or less the same across most western exchanges that I just checked.

23-year history?

Both alternative and mainstream media are epic fails in terms of covering things related to China. The capital control narrative was bullshit to begin with. As is the "CNY strengthening" narrative right now (the rise is insignificant to account for the drop in BTC).

Now, if you buy BTC to bypass CCs you only need them for a limited time as a vehicle. You buy here, dump there - you don't own many for any serious amount of time (accumulation will tend to increase price and also increase the amount of money you lose during the dump), thus the supply equilibrium is not seriously affected to create an increased price.

Plus the average Chinese citizen will never hit the 50k usd limit when CCs kick in. So who is affected really? Big guys. Big guys make their money from owning big businesses. China is an exporting country and one very easy way to get foreign currency while living in an exporting country, is to export stuff. You buy locally with national currency, export the stuff, get paid in $$$ in a foreign bank account and you are ok. Countries which have stuff to sell that the rest of the planet needs, have a very different experience of "capital controls" compared to countries that don't.

There are generally 2 ways to prove it:

1) Burn capital from prior years declared income
2) Show that you've got a loan or donation during the same period that covers the expected difference

If you go route (1) you must have shown this income in past years (so you've already been taxed - you can't escape it)

If you go route (2), both loans and donations have to be declared. Donations are taxed heavily (up to 40%).

So the taxman will get his way.

The origin of the taxes is the problem. If you are alive and breathing, they say you are earning money and have to pay taxes. If you have a car, a home, you are told you earn even more money.

They don't care to speculate where you get your money from, whether it's bitcoin, or something else: It's up to you to pay them from wherever your fictional income comes. And that, in a country with a known rate of >25% unemployment.

Bitcoin is an antidote to many things, but not this kind of taxation scheme.

It's happened since 1/1.

We are now obligated to use electronic transactions for a percentage of our income - whether this income is real or imagined by the government (the government can claim you are making XXXX euro just because you have a car or a house - and tax you based on it).

Bitcoin will not save you from this tax obligation.

Just by being alive, you are considered to earn 3000 euros per year to sustain yourself. "You have to feed yourself, so you are making 3000 euros". And then you are taxed on this fictional income. If you have a car, the government then says, ah, right so now you are 3000+another 3000 due to your car expenses = you are making 6k.... if you also have a house you are considered to be making 8-10k, etc etc. All fictional income, but still very taxable.

Only solution to the above is migration.

I've been thinking about this too. The moment the market becomes predictable ("it will go to 2k-4k-10k"), it's the exact same moment that someone can front run the legitimate non leveraged buyers - making their purchases much harder (while the leveraged trader gets all the money). From that perspective, some uncertainty with periodic dips is a good thing.

Average should include those who overshoot it because they are lazy to calculate it - it's not a requirement.


Pretty much the same as yesterday.

What are you talking about? Blocks are issued normally (13468 transactions in last 6 blocks).


Height   Age   Transactions   Total Sent   Relayed By   Size (kB)
446747   18 minutes   2046   46,210.72 BTC   AntPool   998.19
446746   35 minutes   2463   48,702.46 BTC   F2Pool   999.99
446745   47 minutes   2673   19,541.99 BTC   BitFury   998.12
446744   52 minutes   1880   30,974.79 BTC   F2Pool   999.87
446743   59 minutes   2384   37,938.93 BTC   shawnp0wers   998.22
446742   1 hour 10 minutes   2022   24,366.78 BTC   ViaBTC   999.12

Next time, check this:
https://bitcoinfees.21.co/

Coinbase, banks, etc will all take commissions that will be in the tens of dollars for a 4k USD tx, so it's a crime not to spend 10 extra cents in fees for first priority (hint: there is no standard fee). But I think it'll go upwards in terms of price so in that case you'll be glad it got delayed.


The capital control narrative always sucked. Using BTC as a vehicle does not create such an upward price pressure of 2x (600->1200) in a couple of months.

You buy one place, sell another place, and then loop the process - but the supply/demand equilibrium is ultimately not very affected because you used the BTCs to dump them.

On the other hand, using BTC as a diversified store of value and an investment, in light of a probable devaluation, makes sense.

The chinese are pretty good at understanding resource scarcity (resources / population) because that's what they have been trained to do all their lives, whether it's food, real estate, etc etc.

If you live in a country with 1.4 billion people it's pretty easy to figure out what that means with only 16 million bitcoins around...

Bitstamp 1036.2 BTC-e 1000 Bitfinex 1045.5 Huobi 1082.08/7526

A better milestone would be 1% of gold's marketcap (Gold marketcap=7trn so 1% of that = 70bn)... price would be near $4444

About time

Happy new year

Αποκλειστικα βαση διαγραμματων δε γινεται. Ο λογος ειναι οτι τη πρωτη φορα που θα σου ξεφυγει προς τα πανω ενω εσυ εχεις πουλησει περιμενοντας κινηση προς τα κατω, αν ακολουθεις τυφλα ενα διαγραμμα τοτε πρεπει να περιμενεις αιωνια μεχρι να ξαναπεσει στο σημειο που θες να αγορασεις.

Πχ αν η τιμη παει 200 αλλα εσυ πουλησες και περιμενεις να ξαναγορασεις στο 170, αλλα η τιμη δε ξαναπιασει ποτε 170, τοτε πρακτικα εχεις χασει τα coins σου αλλα και την ανοδο 170->1000.

Οπότε,

1) απλα παραδεχεσαι οτι εκανες λαθος προβλεψη και σε πιανει το FOMO (fear of missing out) ξαναγοραζοντας υψηλοτερα απ'τα 200 που πουλησες οταν το βλεπεις να ανεβαινει και να εχει καλη προοπτικη επιπλεον ανοδου ή
2) το σκεφτεσαι ως εξης "δε γινεται να κανω sell low (200) / buy high (250+) γιατι ετσι δε βγαινουν λεφτα, αρα αστο να παει στο διαλο" και χανεις το τρενο

Για να δουμε αν χτυπησει κανα 1000ρικο πριν το τελος, με το ρυθμο που ανεβαινει....

https://www.wired.com/2007/12/the-voice-of-go/
...
https://www.wired.com/2007/06/darpas_sonic_pr